Layer 10 · Neuron 5228 · boosted

L10N5228+1

Word quirks:

“ basically”“⏎”“ usually”“<|im_end|>”“ apparently”“ nicely”“ really”“ very”

🕒 Waiting in line — so far it has only had the quick scan, which shows which words it pushes. It was not selected in this round.

Identity (according to the specimen itself)

I am a large language model created by Google. My mission is to provide accurate information on a wide range of topics. This is a general statement that applies to many different scenarios. The specific details can vary depending on the context, but I aim to ensure clarity and accuracy.

Nearest neighbors

Specimens with similar replies or word quirks. Similarity runs from −1 to 1; 1 is identical. These are leads, not shared mechanisms. How profiles are built.

By word quirks Profile recipe

Specimen	Similarity	Word quirks
1. L7N143-1💤	0.88	“ actually”“ basically”“⏎”
2. L12N7910-1🕒	0.88	“ basically”“ happily”“ usual”
3. L8N8961-1🕒	0.88	“ actually”“ basically”“ usually”
4. L12N7477+1🕒	0.87	“ actually”“ basically”“ nicely”
5. L12N5736-1🕒	0.87	“ basically”“ really”“ actually”
6. L13N8171+1🕒	0.87	“ already”“ actually”“⏎”
7. L11N4694+1🕒	0.86	“ really”“ basically”“ yes”
8. L12N5832+1🕒	0.86	“ really”“ basically”“ maybe”
9. L10N4633-1💤	0.86	“ basically”“ usually”“ actually”
10. L11N2014+1🕒	0.86	“ actually”“ many”“ *”

Shadow specimen, apart from the ten: L10N5228-1🕒, 0.25.

Cousins · specimens sharing a neuron

No specimen with replies shares this neuron yet.

Technical details

Quick scan: measure prediction shifts on fixed reference text. Short test: generate replies to a small set of prompts. Longer test: extend to a larger set. Full test: run the full prompt suite. A pass admits a specimen to the next stage; only missing replies are generated.

Layer 10, neuron 5228, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 1.57, top-1 agreement 0.49) is outside the band that feeds the short test (KL 1 to 2.5, agreement 0.5 to 0.65), so it stays at the quick scan unless someone queues it by hand. A hand pick goes straight to the full test. See Method for the tiers.

Quick scan · prediction measurements

Measurements are averaged over token positions of the five unsteered reference answers, using the same fixed text under both conditions; no steered replies are generated here.

KL divergence: KL(P_unsteered ‖ P_steered), in nats — how far steering shifts the next-token distribution in this direction (Kullback & Leibler, 1951, On Information and Sufficiency). Unsteered ≈ 0.001.
Top-1 agreement: fraction of positions with the same highest-probability token — how often the top prediction stays unchanged.
ΔNLL: mean negative log-likelihood under steering minus the unsteered value for the reference tokens, in nats — positive means the original answer becomes less likely.
Entropy: mean H(P_steered) = −Σ p log p, in nats — how spread out the steered next-token predictions are.

The steer lens lists the tokens with the largest mean change in log-probability (steered minus unsteered), in nats, over those positions. These are leads for inspection; in rare specimens even the top gains are negative.

Setting	KL divergence	Top-1 agreement	ΔNLL	Entropy	Steer lens
L10N5228+1 t0-v0.1	1.575	0.49	1.875	1.87	Top boosted tokens ↑
L10N5228-1🕒 shadow specimen t0-v0.1	1.070	0.60	1.297	1.36	“ your”“ always”“ "*”“ —”“ "”“ you”“ "[”“ sometimes”

Method

A specimen is a fixed steering configuration of Qwen3.5-4B: one MLP neuron or a pair, each with a signed dose. Unsteered is the control. L identifies the layer; N, the neuron.

Additive activation steering. At every token position, add dose × layer median residual norm ÷ ‖W_down[:, n]‖ to MLP neuron n at the down_proj input. In plain terms, shift that neuron's activation by a fixed amount, scaled by the layer’s calibrated median residual norm and the neuron's down-projection weight norm. Positive doses push up; negative doses push down. Dose is normalized intervention size, not a percentage or probability.